Bumjun Kim

selected publications

(*) denotes equal contribution, (†) denotes corresponding author.

1. ICML

   Dependency-Aware Parallel Decoding via Attention for Diffusion LLMs

   Bumjun Kim^*, Dongjae Jeon^*, Moongyu Jeon, and Albert No^†

   In the International Conference on Machine Learning, 2026.

   PDF

   Parallel decoding for diffusion LLMs using attention-derived dependency structure.

2. ICLR

   Rainbow Padding: Mitigating Early Termination in Instruction-Tuned Diffusion LLMs

   Bumjun Kim^*, Dongjae Jeon^*, Dueun Kim^*, Wonje Jeung, and Albert No^†

   In the International Conference on Learning Representations, 2026.

   OpenReview

   A method for reducing early termination in instruction-tuned diffusion language models.

3. CVPRFindings

   Memorization In Stable Diffusion Is Unexpectedly Driven by CLIP Embeddings

   Bumjun Kim and Albert No^†

   In the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2026 Findings.

   PDF

   An analysis of Stable Diffusion memorization that identifies CLIP embeddings as a key driver.

4. arXiv

   Understanding the Reversal Curse Mitigation in Masked Diffusion Models through Attention and Training Dynamics

   Sangwoo Shin, Bumjun Kim, Kyelim Lee, Moongyu Jeon, and Albert No^†

   arXiv, 2026.

   PDF

   A study of attention and training dynamics behind reversal-curse mitigation in masked diffusion models.